CBSE Launched an AI and Computational Thinking Curriculum for Classes 3 to 8 on 1 April 2026, and 5.17 Lakh Learners Had Enrolled in the Government's SOAR AI Courses by 6 August 2026

CBSE’s curriculum on computational thinking and artificial intelligence for Classes 3 to 8 began with the 2026-27 session. Education Minister Dharmendra Pradhan launched it at Vigyan Bhawan in New Delhi on 1 April 2026, and CBSE’s circular of the same date, Acad-15/2026, asked affiliated schools to put it into effect. Computational thinking is the habit of breaking a problem into small steps and spotting patterns, and the curriculum treats it as the base on which AI is taught. Everything below is as of 10 October 2026.

The decision and the dates on record

The Ministry of Education announced the plan in a PIB release of 30 October 2025. It followed a stakeholder consultation on 29 October 2025 with CBSE, NCERT, Kendriya Vidyalaya Sangathan (KVS), Navodaya Vidyalaya Samiti (NVS) and outside experts. CBSE had formed an expert committee chaired by Prof. Karthik Raman of IIT Madras. School Education Secretary Sanjay Kumar said AI education should be treated as a basic universal skill tied to the subject The World Around Us.

At the launch on 1 April 2026, Pradhan described the curriculum as structured AI education brought into schools at scale, under the line “AI for Education, AI in Education”. Minister of State Jayant Chaudhary, Secretaries Sanjay Kumar and Vineet Joshi, CBSE Chairman Rahul Singh and NCERT Director Dinesh Prakash Saklani were present, according to PIB. The circular names logical reasoning, problem-solving, pattern recognition and algorithmic thinking as the skills to build, with digital literacy and ethical use of technology alongside.

50 hours in Classes 3 to 5, 100 in Classes 6 to 8

The CBSE curriculum document sets a suggested time of 50 hours a year for Classes 3 to 5 and 100 hours a year for Classes 6 to 8. It says the figures are meant to keep the existing load unchanged and that schools may go beyond them. In Classes 3 to 5 the work sits inside Mathematics and The World Around Us, through resource books that add computational-thinking questions to the chapters already taught. A Class 3 pupil, for example, is expected to follow step-by-step rules for movement on grids and to spot patterns where one or two things change from term to term.

Assessment in Classes 3 to 5 is by written tests with puzzles, group activities such as treasure hunts, and a teacher’s observation journal. CBSE’s circular of 9 April 2026 says the framework aims to create AI-literate learners by 2030.

What Classes 6, 7 and 8 learn about AI

The AI part of the middle stage is 20 hours in each of Classes 6, 7 and 8, split into four units of 5 hours. The other hours go to computational-thinking worksheets and interdisciplinary projects.

• Class 6: what AI is and how it differs from automation, kinds of data (numbers, text, images, sound), simple patterns and decisions, and online safety, passwords and digital footprints.
• Class 7: classification, regression and clustering tried on a small dataset with AI tools, computer vision and language processing, AI in healthcare, education and transport, and bias.
• Class 8: the stages of an AI project (define the problem, collect data, test tools, reflect and improve), hands-on work with no-code tools such as image classifiers and chatbots, fairness in data, privacy and misinformation.

The curriculum assigns computational-thinking resources to subject teachers and AI literacy to computer teachers, who also assess the combined projects.

Handbooks, teachers and the summer training

Circular Acad-18/2026, dated 9 April 2026, announced student and teacher resource books for Classes 3 to 8, and CBSE’s CT and AI page lists a student handbook and a teacher handbook for each of the six grades. The circular describes them as companions to the NCERT Mathematics textbooks, following the same chapter order. The 30 October 2025 release had set December 2025 as the date for handbooks and digital resources.

Akashvani reported on 12 May 2026 that Sanjay Kumar said teachers would be trained during the summer break and that teaching would begin when schools reopened. He added that NCERT is developing AI textbooks for Classes 9 to 12. For senior classes, the same 9 April circular says other AI courses run until 2025-26 are discontinued for Class IX from 2026-27, Class X continues its earlier scheme, and NCERT will supply computational-thinking and AI modules for Classes 9 to 12 as internal-assessment modules. AI continues as a skill subject in Classes XI and XII: CBSE’s notification of 24 June 2026 schedules online programmes with IBM for those teachers, from 1 July to 5 November 2026.

SOAR and the other skilling programmes

The Ministry of Skill Development and Entrepreneurship launched Skilling for AI Readiness (SOAR) on 22 July 2025. Its first four courses are three 15-hour student modules (AI to be Aware, AI to Acquire, AI to Aspire) for Classes 6 to 12 and a 45-hour module for teachers called AI for Educators. All are online and self-paced on the Skill India Digital Hub. The programme later grew to 50 AI-related qualifications, and PIB’s note of 12 August 2026 reports 5,17,477 learners enrolled and 1,00,664 certified under SOAR courses as on 6 August 2026. Those totals cover every SOAR course, not only the school ones.

PIB’s release of 9 March 2026 gives state-wise data, as of 23 February 2026, for the four first-phase courses. Added up by tuput, it shows 2,30,901 enrolments and 29,771 certificates, of which AI for Educators accounts for 40,627 enrolments and 4,178 certificates. The programme is open to learners, educators and citizens, so the table does not separate school pupils. A separate MeitY programme, YUVAi, has run since November 2022 and trains students of Classes 8 to 12 to apply AI across eight themes, among them agriculture, health and education, according to PIB’s backgrounder of 10 February 2026. The money and compute behind the wider IndiaAI Mission are set out in an earlier tuput piece.

A Rs 500 crore centre and the school infrastructure in the Budget

The Union Budget of 1 February 2025 announced a Centre of Excellence in AI for education with a total outlay of Rs 500 crore. The same Budget set out 50,000 Atal Tinkering Labs in government schools over five years and broadband connectivity for all government secondary schools under the BharatNet project. Pradhan launched the Bodhan AI Centre of Excellence at IIT Madras on 12 February 2026, at the Bharat Bodhan AI Conclave in Bharat Mandapam, Delhi. PIB describes the conclave as the start of the Bharat EduAI Stack, an open architecture for building and scaling AI learning tools, and quotes Pradhan on building a digital public infrastructure (DPI) for education, the model behind UPI and Aadhaar that tuput covers in a piece on India’s DPI abroad. Speech and translation scores for Indian languages, including work from Bodhan, are covered in another tuput piece.

What Punjab, Andhra Pradesh and Karnataka have announced

Careers360 reported on 23 July 2026 that Punjab’s Education Minister Harjot Singh Bains had announced AI as a mainstream subject for Classes 1 to 12 under a project called Sarbat.AI. The state calls itself the first to do so, and the claim rests on the government’s own statement. The first phase covers Classes 8 to 12 inside computer science and the second covers Classes 1 to 7. The report puts the reach at 25,172 schools and nearly 31.5 lakh students, with training planned for more than 2 lakh teachers.

Speaking after the PanIIT Summit in Vijayawada, in a report published on 4 October 2026, Andhra Pradesh Chief Minister N. Chandrababu Naidu said AI would be taught from Class 6 starting next academic year, according to The Week. He said the state would ask IIT Hyderabad to write the high-school syllabus, which means the syllabus is still to be written. PTI reported on 5 October 2026 that Karnataka Chief Minister D K Shivakumar directed officials to explore AI training in government schools from Class 6, with the IT department and private partners, and the statement said training has already started for Class 9 students in government schools.

Computers and internet in schools, and the next stage

The Ministry’s UDISE+ report for 2025-26, released on 7 July 2026, shows schools with computer access at 69.9 per cent, up from 64.7 per cent in 2024-25 and 47.7 per cent in 2022-23. Internet access is at 67.4 per cent, up from 63.5 per cent and 49.7 per cent. The report counts 1,02,73,020 teachers (about 1.03 crore).

In its Rajya Sabha reply of 5 August 2026, the Ministry told Parliament that CBSE schools must keep the infrastructure the curriculum lists for AI and must maintain digital infrastructure under the Board’s affiliation rules. It added that education is on the Concurrent List and most schools sit under State and Union Territory governments, so AI and robotics laboratories in those schools are the states’ responsibility. The documents opened for this piece give no count of teachers trained so far for the new curriculum. The reply also says the curriculum for Grades 9 to 12 in AI and allied subjects has been finalised.